pubs.acs.org/Langmuir © 2009 American Chemical Society
Controlled Oxidation, Biofunctionalization, and Patterning of Alkyl Monolayers on Silicon and Silicon Nitride Surfaces using Plasma Treatment Michel Rosso,†,‡ Marcel Giesbers,† Karin Schroen,‡ and Han Zuilhof *,† †
Laboratory of Organic Chemistry, Wageningen University, Dreijenplein 8, 6703 HB Wageningen, The Netherlands, and ‡Laboratory of Food and Bioprocess Engineering, Wageningen University, Bomenweg 2, 6703 HD Wageningen, The Netherlands Received June 27, 2009 A new method is presented for the fast and reproducible functionalization of silicon and silicon nitride surfaces coated with covalently attached alkyl monolayers. After formation of a methyl-terminated 1-hexadecyl monolayer on Hterminated Si(100) and Si(111) surfaces, short plasma treatments (1-3 s) are sufficient to create oxidized functionalities without damaging the underlying oxide-free silicon. The new functional groups can, e.g., be derivatized using the reaction of surface aldehyde groups with primary amines to form imine bonds. In this way, plasma-treated monolayers on silicon or silicon nitride surfaces were successfully coated with nanoparticles, or proteins such as avidin. In addition, we demonstrate the possibility of micropatterning, using a soft contact mask during the plasma treatment. Using water contact angle measurements, ellipsometry, XPS, IRRAS, AFM, and reflectometry, proof of principle is demonstrated of a yet unexplored way to form patterned alkyl monolayers on oxide-free silicon surfaces.
Introduction Plasma treatments of organic materials have been widely studied to prepare modified surfaces for organic membranes1 or materials of biotechnological and biomedical interests.2-4 Indeed, a short exposure of organic surfaces, usually polymers, to plasma can create directly new surface functionalities. The gas present in the plasma chamber determines the obtained functionalization:5,6 oxygen or water plasma leads to the oxidation of surfaces and to the formation of polar surface groups (-OH, CdO, O-CdO), whereas exposure to ammonia, for instance, will mainly yield surface amine groups (-NH2). In most cases, these treatments produce surfaces with a higher biocompatibilty,7-9 but another interesting application involves the subsequent functionalization of these surfaces with bioactive molecules for specific recognition and sensing.10 While plasma treatments yield lower densities of surface functional groups than classical chemical reactions, this can be compensated by the size of the subsequently grafted moieties, whether they consist of biomolecules2 (enzymes, antibodies, DNA, etc.) or polymer brushes.11,12 *E-mail:
[email protected]. (1) Durand, J.; Rouessac, V.; Roualdes, S. Ann. Chim.-Sci. Mater. 2007, 32, 141–158. (2) Siow, K. S.; Britcher, L.; Kumar, S.; Griesser, H. J. Plasma Process. Polym. 2006, 3, 392–418. (3) Denes, F. Trends Polym. Sci. 1997, 5, 23–31. (4) Uyama, Y.; Kato, K.; Ikada, Y. Adv. Polym. Sci. 1998, 137, 1–39. (5) Poncin-Epaillard, F.; Legeay, G. J. Biomater. Sci. Polymer Ed. 2003, 14, 1005–1028. (6) Vandencasteele, N.; Merche, D.; Reniers, F. Surf. Interface Anal. 2006, 38, 526–530. (7) Giroux, T. A.; Cooper, S. L. J. Colloid Interface Sci. 1990, 139, 351–362. (8) Ulbricht, M.; Belfort, G. J. Membr. Sci. 1996, 111, 193–215. (9) Ulbricht, M.; Belfort, G. J. Appl. Polym. Sci. 1995, 56, 325–343. (10) Steffen, H. J.; Schmidt, J.; Gonzalez-Elipe, A. Surf. Interface Anal. 2000, 29, 386–391. (11) Kingshott, P.; Thissen, H.; Griesser, H. J. Biomaterials 2002, 23, 2043–2056. (12) Kingshott, P.; McArthur, S.; Thissen, H.; Castner, D. G.; Griesser, H. J. Biomaterials 2002, 23, 4775–4785. (13) Raacke, J.; Giza, M.; Grundmeier, G. Surf. Coat. Technol. 2005, 200, 280– 283. (14) Elms, F. M.; George, G. A. Polym. Adv. Technol. 1998, 9, 31–37. (15) Dai, X. J.; Elms, F. M.; George, G. A. J. Appl. Polym. Sci. 2001, 80, 1461– 1469. (16) Wagner, A. J.; Wolfe, G. M.; Fairbrother, D. H. J. Chem. Phys. 2004, 120, 3799–3810.
866 DOI: 10.1021/la9023103
Beside the work carried out on polymer surfaces, several groups have also studied the effects of oxygen plasma,13-15 as well as atomic oxygen16,17 or ion beams18,19 on organic thiol monolayers on gold. In comparison, very little has been done concerning the further functionalization of plasma-treated monolayers. This is somewhat surprising, as plasma treatment provides an easy and fast activation of chemically inert organic monolayers (e.g., methyl-terminated), with potential applications in biosensing. Recently, carbon dioxide20 and oxygen21,22 plasma treatments were used to functionalize alkylsilane monolayers on glass and silica surfaces, and one of the latter works demonstrated the attachment of antibodies.22 In this work, we extended the plasma functionalization to alkyl monolayers on oxide-free silicon, produced from the reaction of alkenes with hydrogen-terminated silicon surfaces and with HFetched silicon-enriched silicon nitride (Si3.9N4). These high-quality alkyl monolayers, prepared by thermal or photochemical reaction of alkenes with hydrogen-terminated silicon surfaces,23-34 have been studied for their high potential in sensing and nanotechnology (17) Torres, J.; Perry, C. C.; Bransfield, S. J.; Fairbrother, D. H. J. Phys. Chem. B 2002, 106, 6265–6272. (18) Qin, X. D.; Tzvetkov, T.; Jacobs, D. C. J. Phys. Chem. A 2006, 110, 1408– 1415. (19) Qin, X.; Tzvetkov, T.; Jacobs, D. C. Nucl. Instrum. Methods Phys. Res., Sect. B 2003, 203, 130–135. (20) Delorme, N.; Bardeau, J. F.; Bulou, A.; Poncin-Epaillard, F. Thin Solid Films 2006, 496, 612–618. (21) Unger, W. E. S.; Lippitz, A.; Gross, T.; Friedrich, J. F.; Woll, C.; Nick, L. Langmuir 1999, 15, 1161–1166. (22) Xue, C. Y.; Yang, K. L. Langmuir 2007, 23, 5831–5835. (23) Boukherroub, R.; Morin, S.; Bensebaa, F.; Wayner, D. D. M. Langmuir 1999, 15, 3831–3835. (24) Buriak, J. M. Chem. Rev. 2002, 102, 1271–1308. (25) Eves, B. J.; Sun, Q. Y.; Lopinski, G. P.; Zuilhof, H. J. Am. Chem. Soc. 2004, 126, 14318–14319. (26) Linford, M. R.; Fenter, P.; Eisenberger, P. M.; Chidsey, C. E. D. J. Am. Chem. Soc. 1995, 117, 3145–3155. (27) Shirahata, N.; Hozumi, A.; Yonezawa, T. Chem. Rec. 2005, 5, 145–159. (28) Sieval, A. B.; Linke, R.; Zuilhof, H.; Sudh€olter, E. J. R. Adv. Mater. 2000, 12, 1457–1460. (29) Sieval, A. B.; Opitz, R.; Maas, H. P. A.; Schoeman, M. G.; Meijer, G.; Vergeldt, F. J.; Zuilhof, H.; Sudh€olter, E. J. R. Langmuir 2000, 16, 10359–10368. (30) Sun, Q.-Y.; de Smet, L. C. P. M.; van Lagen, B.; Giesbers, M.; Thune, P. C.; van Engelenburg, J.; de Wolf, F. A.; Zuilhof, H.; Sudh€olter, E. J. R. J. Am. Chem. Soc. 2005, 127, 2514–2523.
Published on Web 09/03/2009
Langmuir 2010, 26(2), 866–872
Rosso et al.
applications,35-40 due to the possibility to attach biomolecules directly onto the semiconductor surfaces.41-49 Importantly, these densely packed monolayers possess a much higher stability than thiol monolayers on gold or alkylsiloxane monolayers on oxide surfaces, because of the formation of strong Si-C bonds,24,26,50 which makes them more robust for applications in aqueous solutions.51 Within the alkenes available for the formation of methylterminated alkyl monolayers, methyl-terminated, monofunctional alkenes are probably the most convenient to use: they are commercially available at low prices with any specified length, and if so desired, their further purification is easy (distillation). In addition, their grafting conditions are very flexible, usually in the liquid state, while 1-alkenes are both heat-resistant and UV-resistant (>250 nm). Once these stable monolayers are formed, a short plasma treatment (0.5-2 s) is able to form oxidized functionalities within the top few A˚ngstroms of the surface, while the underlying alkyl chains retain their initial packing and insulation properties of the inorganic substrate. In addition, we show the easy formation of coatings of gold nanoparticles and functional proteins. In this latter case, the attachment of avidin was used, and the specific interaction with biotin-labeled bovine serum albumin (BSA) was monitored on the surface with reflectometry. Further, we demonstrate the feasibility of surface patterning, using a soft contact mask during the plasma oxidation. To our knowledge, this is the first attempt to functionalize alkyl monolayers on oxide-free silicon or silicon nitride using plasma. Besides the fundamental interest raised by (31) Sun, Q.-Y.; de Smet, L. C. P. M.; van Lagen, B.; Wright, A.; Zuilhof, H.; Sudh€olter, E. J. R. Angew. Chem., Int. Ed. 2004, 43, 1352–1355. (32) Wayner, D. D. M.; Wolkow, R. A. J. Chem. Soc., Perkin Trans. 2 2002, 23–34. (33) Boukherroub, R. Curr. Opin. Solid State Mater. Sci. 2005, 9, 66–72. (34) Sieval, A. B.; Demirel, A. L.; Nissink, J. W. M.; Linford, M. R.; van der Maas, J. H.; de Jeu, W. H.; Zuilhof, H.; Sudh€olter, E. J. R. Langmuir 1998, 14, 1759–1768. (35) Lasseter, T. L.; Clare, B. H.; Abbott, N. L.; Hamers, R. J. J. Am. Chem. Soc. 2004, 126, 10220–10221. (36) Aswal, D. K.; Lenfant, S.; Guerin, D.; Yakhmi, J. V.; Vuillaume, D. Anal. Chim. Acta 2006, 568, 84–108. (37) Zhao, J. W.; Uosaki, K. J. Phys. Chem. B 2004, 108, 17129–17135. (38) Faber, E. J.; Sparreboom, W.; Groeneveld, W.; de Smet, L.; Bomer, J.; Olthuis, W.; Zuilhof, H.; Sudholter, E. J. R.; Bergveld, P.; van den Berg, A. ChemPhysChem 2007, 8, 101–112. (39) Faber, E. J.; de Smet, L. C. P. M.; Olthuis, W.; Zuilhof, H.; Sudholter, E. J. R.; Bergveld, P.; van den Berg, A. ChemPhysChem 2005, 6, 2153–2166. (40) Yaffe, O.; Scheres, L.; Puniredd, S.; Stein, N.; Biller, A.; Har-Lavan, R.; Shpaisman, H.; Zuilhof, H.; Haick, H.; Cahen, D.; Vilan, A. Nano Lett. 2009, 9, 2390–2394. (41) Letant, S. E.; Hart, B. R.; Van Buuren, A. W.; Terminello, L. J. Nat. Mater. 2003, 2, 391–396. (42) Liao, W.; Wei, F.; Qian, M. X.; Zhao, X. S. Sens. Actuators, B 2004, 101, 361–367. (43) Voicu, R.; Boukherroub, R.; Bartzoka, V.; Ward, T.; Wojtyk, J. T. C.; Wayner, D. D. M. Langmuir 2004, 20, 11713–11720. (44) de Smet, L. C. P. M.; Stork, G. A.; Hurenkamp, G. H. F.; Sun, Q. Y.; Topal, H.; Vronen, P. J. E.; Sieval, A. B.; Wright, A.; Visser, G. M.; Zuilhof, H.; Sudh€olter, E. J. R. J. Am. Chem. Soc. 2003, 125, 13916–13917. (45) Lin, Z.; Strother, T.; Cai, W.; Cao, X. P.; Smith, L. M.; Hamers, R. J. Langmuir 2002, 18, 788–796. (46) Strother, T.; Hamers, R. J.; Smith, L. M. Nucleic Acids Res. 2000, 28, 3535– 3541. (47) Hong, Q.; Rogero, C.; Lakey, J. H.; Connolly, B. A.; Houlton, A.; Horrocks, B. R. Analyst 2009, 134, 593–601. (48) (a) Yang, M.; Teeuwen, R. L. M.; Giesbers, M.; Baggerman, J.; Arafat, A.; de Wolf, F. A.; van Hest, J. C. M.; Zuilhof, H. Langmuir 2008, 24, 7931–7938. (b) Sieval, A. B.; Huisman, C. L.; Schonecker, A.; Schuurmans, F. M.; van der Heide, A. S. H.; Goossens, A.; Sinke, W. C.; Zuilhof, H.; Sudholter, E. J. R. J. Phys. Chem. B 2003, 107, 6846–6852. (49) Bocking, T.; Killan, K. A.; Gaus, K.; Gooding, J. J. Langmuir 2006, 22, 3494–3496. (50) Bateman, J. E.; Eagling, R. D.; Worrall, D. R.; Horrocks, B. R.; Houlton, A. Angew. Chem., Int. Ed. 1998, 37, 2683–2685. (51) Sano, H.; Maeda, H.; Ichii, T.; Murase, K.; Noda, K.; Matsushige, K.; Sugimura, H. Langmuir 2009, in press, DOI: 10.1021/la804080g.
Langmuir 2010, 26(2), 866–872
Article
the behavior of monolayers submitted to the plasma treatment, this technique provides a very fast (a few seconds) and reproducible grafting and dry patterning method for silicon surfaces.
Materials and Methods 1-Hexadecene (99%) was purchased from Sigma-Aldrich and distilled twice under reduced pressure before use. m-(Trifluoromethyl)benzylamine (TFBA) (97%), cysteamine (2-aminoethanthiol), dimethyl sulfoxide (DMSO, spectrophotometric grade, 99.9%, dried over molecular sieves), and N-methyl-2-pyrrolidone (NMP, anhydrous, 99.5%) were purchased from Sigma-Aldrich and used as received for surface reactions. Citrate-terminated gold nanoparticles (Au NPs; diameter 15 nm) were purchased from Aurion BV, Wageningen, The Netherlands. Biotin hydrazide (98%), bovine serum albumin (BSA: fraction V, min 96% lyophilized powder), biotin-labeled BSA (lyophilized powder, 80%, extent of labeling: 8-16 mol biotin per mol BSA), and avidin (recombinant protein from egg white, expressed in corn, ∼ 12 units/mg of protein) were purchased from Sigma. Silicon Surface Modification. 1-Hexadecyl-terminated silicon surfaces were obtained via a thermal modification of silicon wafers in neat alkenes.52 Silicon wafers of Si(100) ( 0.5° and Si(111) ( 0.5°, (both N-type Si, resistivity 1-5 Ω.cm; Siltronix, France) were first cleaned by sonication in acetone, followed by oxidation in air-based plasma for 5 min. Si(100) substrates were then H-terminated using etching in a 2.5% aqueous HF solution for 2 min, whereas Si(111) substrates were etched in an argon-saturated 40% aqueous solution of NH4F for 15 min and subsequently rinsed with ultrapure water. Right after etching, substrates were dried with an argon flow, and dipped into argon-saturated neat 1-hexadecene at 200 °C to react for 4 h. After reaction, the modified substrates were rinsed thoroughly with petroleum ether and acetone and finally sonicated in acetone. Silicon Nitride Surface Modification. Si-rich silicon nitride (SixN4, x ≈ 3.9) deposited by chemical vapor deposition (CVD) on Si(100) (thickness of 147 nm) was obtained from Lionix B.V., The Netherlands. 1-Hexadecyl-terminated silicon nitride substrates were obtained according to our previous work:53 SixN4 samples (with sizes of 1 1 cm for X-ray photoelectron spectroscopy (XPS) or 4 0.75 cm for reflectometry) were cleaned by sonication in acetone, followed by oxidation in an air-based plasma for 15 min. The oxidized samples were then etched with a 2.5% aqueous HF solution for 2 min and dried in an argon flow. They were then immediately dipped into argon-saturated neat alkenes in a fused silica flask. After 30 more min under argon flow, a UV pen lamp (254 nm, low-pressure mercury vapor, double-bore lamp from Jelight Company, California, U.S.A.) was placed 4 mm above the SixN4 surface and the sample was irradiated for 24 h. Afterward, samples were removed and rinsed several times with petroleum ether and acetone and sonicated in the same solvents. Plasma Oxidation and Derivatization. Monolayer-coated substrates were oxidized using a Harricks plasma cleaner/sterilizer (RF power 7.2 W), under a pressure of 5 10-2 mbar, with an air flow of about 25 L/h. After evacuation of the plasma chamber, the plasma was turned on for the desired time (0.5-3 s). Oxidized substrates were then directly used for subsequent treatments or analysis. (52) Scheres, L.; Arafat, A.; Zuilhof, H. Langmuir 2007, 23, 8343–8346. (53) Rosso, M.; de Jong, E.; Giesbers, M.; Fokkink, R. G.; Norde, W.; Schroen, K.; Zuilhof, H., in preparation.
DOI: 10.1021/la9023103
867
Article
Amine attachment: Plasma-treated surfaces were dipped into 0.01 M solutions of TFBA in dry DMSO for 24 h. After reaction, samples were rinsed with DMSO and acetone, sonicated in acetone, and dried in an argon flow. Attachment of Au NPs: Plasma-treated surfaces were reacted with 10-2 M solutions of cysteamine in NMP for 24 h. The obtained thiol-terminated surfaces were then reacted with Au NPs in water for 2 h. After rinsing with water and acetone, samples were sonicated in acetone and dried. Avidin attachment: The plasma-oxidized surfaces were dipped for 24 h in a 10-2 M solution of biotin hydrazide in DMSO at room temperature for 24 h, followed by a 1 h incubation in 1 mL of a 1 mg/L solution of avidin in phosphate buffered saline (PBS) buffer (pH 6.75). After attachment of avidin, the substrates were incubated in a 0.1 g/L solution of unlabeled BSA for 1 h, to passivate the remaining free area of the surface and thus avoid nonspecific adsorption of protein during the reflectometry measurements. Static Water Contact Angle Measurements. The wetting properties of modified surfaces were characterized by automated static water contact angle measurements performed with an Erma Contact Angle Meter G-1 (volume of the drop of demineralized water =3.5 μL). Ellipsometry. Thickness measurements were performed with a computer-controlled null ellipsometer (Sentech SE-400) using a He-Ne laser (λ =632.8 nm) and an incident angle of 70°. The mode “polarizer þ retarder, aperture, strict” was used. The layer thickness was determined using a three-layer model in the ellipsometry software from Sentech. Fourier Transform Infrared Reflection Absorption Spectroscopy (FT-IRRAS). Spectra were measured with a Bruker Tensor 27 FT-IR spectrometer, using a commercial variableangle reflection unit (Auto Seagull, Harrick Scientific). A Harrick grid polarizer was installed in front of the detector and was used to measure spectra with p-polarized (parallel) radiation with respect to the plane of incidence at the sample surface. Single-channel transmittance spectra were collected using a spectral resolution of 4 cm-1, using 1024 scans in each measurement. The optimal angle for data collection was found to be 68° for all the surfaces studied. All the reported measurements were performed at this angle. Raw data files were divided by data recorded on a plasma-oxidized reference surface, to give the reported spectra. X-ray Photoelectron Spectroscopy (XPS). The XPS analysis was performed using a JPS-9200 Photoelectron Spectrometer (JEOL, Japan). The high-resolution spectra were obtained under UHV conditions using a monochromatic Al KR X-ray radiation at 12 kV and 25 mA, using an analyzer pass energy of 10 eV and a takeoff angle of 1° relative to the surface normal. All high-resolution spectra were corrected with a linear background before fitting. Binding energies were calibrated at 285.0 eV for the C1s peak corresponding to carbon in alkyl chains. To facilitate the comparison between samples, XPS intensities measured on Si were normalized to the intensity of the Si2p signal. Atomic Force Microscopy (AFM). Images were obtained with an MFP-3D AFM from Asylum Research (Santa Barbara, CA). Imaging was performed in AC mode in air using OMCLAC240 silicon cantilevers (Olympus Corporation, Japan). Reflectometry. The setup used for reflectometry measurements has been described.53 In short, a monochromatic light beam (He-Ne laser, λ = 632.8 nm) is linearly polarized and passes a 45° glass prism. This beam arrives at the interface with an angle of incidence of 66° for the solvent/substrate interface. 868 DOI: 10.1021/la9023103
Rosso et al.
Figure 1. (a) Water contact angles and (b) ellipsometric thicknesses measured on hexadecyl-terminated Si(111) and Si(100) surfaces, before and after different times of plasma treatments.
After reflection at the interface and refraction in the prism, the beam is split into its p- and s-polarized components relative to the plane of incidence by means of a beam splitter. Both components are separately detected by two photodiodes, and the ratio between the intensity of the parallel and perpendicular components is the output signal S (S =Ip/Is). Combined with a stagnation point flow cell, the setup allows the introduction of buffer or protein solutions, to study homogeneous adsorption on surfaces under diffusion-controlled conditions. Strips of SixN4-coated silicon wafer (typical size of 4 0.75 cm) were coated with a monolayer on one end (about one-half of the sample length), while the other end was used to hold the strip in the measuring cell of the reflectometer. The BSA and biotinlabeled BSA (0.1 mg/L) were freshly prepared in PBS buffer (pH 6.75, ionic strength 0.08 M). All reflectometry experiments were performed at 23 °C. Before measurements were made, surfaces were incubated 1 h in buffer to avoid artifacts due to initial surface wetting. After placing the samples in the reflectometer, the buffer solution was injected until the output signal was nearly constant: fluctuations of less than 0.001 V over 2 min were considered satisfactory. The adsorbed amounts of protein were calculated according to the method described in the Supporting Information.
Results and Discussion Formation and Plasma Treatment of 1-Hexadecyl Monolayers on Silicon. After 4 h of reaction in neat 1-hexadecene at 200 °C, the silicon substrates were coated with a hydrophobic hexadecyl monolayer. Water contact angles of 109 ( 1° and 111 ( 1° were measured on Si(100) and Si(111) surfaces, respectively, in line with literature data.52 No significant difference in layer thickness was measured with ellipsometry, giving monolayers on both surface orientations a thickness of 1.7 ( 0.1 nm. When submitted to short plasma treatments, these hexadecyl monolayers on Si(100) and Si(111) surfaces showed similar behavior, when the reaction was monitored through water contact angle and monolayer thickness (See Figure 1): in the first 3 s of treatment, the water contact angles decreased from g109° to 0°, indicating a complete surface coverage of polar oxidized species. Simultaneously, the monolayer thickness decreased from the initial 1.7 ( 0.1 nm down to about 1.3 ( 0.2 nm after 3 s. Shorter treatment times resulted in intermediate contact angles in a very reproducible way, indicating a progressive oxidation of surfaces. In particular, monolayers treated for 0.5 s (water contact angles: 89 ( 3°) were as thick as the initial 1-hexadecyl coatings (1.7 ( 0.1 nm) but less hydrophobic. At this stage, the oxidation of surfaces started, but no significant part of the monolayer was removed. After 1 s of treatment, the further decrease in contact angle (64 ( 3°) was accompanied by a small decrease in thickness Langmuir 2010, 26(2), 866–872
Rosso et al.
Article
Figure 3. (a) Relative proportions of the C1s XPS signals at different times of plasma treatment (values for C-Si are omitted for clarity). (b) Normalized F1s intensity after coupling of oxidized surfaces with TFBA to form the fluorinated imine.
Figure 4. Si2p region of the XPS data measured on 1-hexadecyl monolayers on (a) Si(100) and (b) Si(111) at different times of plasma treatment.
Figure 2. C1s region of XPS data measured on plasma-treated hexadecyl monolayers on Si(111) substrates after various reaction times.
of about 0.2 nm for both Si(100) and Si(111), indicating the oxidative fractionation of alkyl chains by the plasma. After 3 s of treatment, about a third of the monolayer thickness was removed. The XPS analysis at the different stages of the plasma treatment gives more insight into the newly created surface functionalities. The C1s region of the high-resolution XPS data measured at 0-3 s of plasma treatment reveals clearly the appearance of new oxygenated species (see Figure 2 for the case of Si(111)). The initial 1-hexadecyl monolayer showed a main peak at 285.0 eV, characteristic for alkyl chains (-CHn-), and a small peak at 283.7 eV due to the terminal carbon bound to silicon (C-Si).52 After plasma treatment, a decrease in the intensity of the CHn peak at 285.0 eV is associated with an increase in the intensity of peaks at higher binding energies. This is in agreement with previous XPS results for thiol monolayers on gold,15,54 and for organosilane monolayers on oxidized surfaces.21,22 Indeed, these results are typical for plasma-treated hydrocarbon surfaces, where the main reactive component is atomic oxygen, which first abstracts hydrogen atoms from the surfaces.55 In turn, the remaining carbon radicals react with oxygen to yield the polar hydroxyl, carbonyl, and carboxyl groups observed with XPS. Interestingly, the total C1s intensity did not decrease after 0.5 s of treatment, which implies that the top methyl groups are only partly oxidized and not yet removed. This confirms the conclusions already drawn from the ellipsometric measurements shown in Figure 1b. Despite the complexity of the plasma-induced (54) Elms, F. M.; George, G. A. Polymer Adv. Tech. 1998, 9, 31–37. (55) Dai, X. J.; Hamberger, S. M.; Bean, R. A. Aust. J. Phys. 1995, 48, 939–951.
Langmuir 2010, 26(2), 866–872
oxidation reactions, which involve in particular the formation of highly reactive oxygen atoms,55 the oxygen-bound carbons can be grouped under three main contributions at 286.6 eV (C-O), 288.3 eV (CdO), and 289.7 eV (O-CdO). Figure 3a gives an overview of the contributions of these different carbon species. Carbon atoms with a single bond to oxygen (C-O) seem the most favored throughout the whole treatment. After 1 s of oxidation, CdO bonds start to represent a significant part of the surface functionalities, most probably in the form of aldehyde groups that can react with primary amines to form imine bonds.22 The proportion of these CdO bonds on the surface did not increase much for plasma treatments longer than 1 s. This was confirmed by coupling these oxidized substrates with metatrifluoromethylbenzylamine (TFBA; m-CF3-C6H4-CH2NH2) that reacts rapidly with carbonyl moieties, especially of aldehyde groups: samples treated for 1 s with plasma gave a maximum value of the normalized intensity of the F1s peak measured on these samples at 689.0 eV. This maximum value, after correction for the different sensitivity factors of the C1s and F1s peaks, indicates that 35-40% of the CdO groups introduced after plasma treatment have reacted. The remaining part probably consists of aldehyde groups too hindered to react with the bulky TFBA molecules, or functionalities other than aldehydes that were also counted in the peak at 288.3 eV but with no or little reactivity toward amines under these conditions. Similar observations were made from the C1s region of XPS data measured on Si(100) surfaces (see Figure S2 of the Supporting Information). Unlike the C1s region of XPS data, the Si2p spectra measured on plasma-treated samples reveal an important difference between the Si(100) and Si(111) surfaces (see Figure 4): only for monolayers on Si(100) can we see the growth of a peak between 102.5 and 103.0 eV, corresponding to the formation of SiO2 at the monolayer-substrate interface. After 3 s of treatment, this SiO2 peak corresponds to 5% of the total Si2p signal. In contrast to this DOI: 10.1021/la9023103
869
Article
Figure 5. IRRAS data in the regions of (a) CHn and (b) CdO stretching vibrations, measured on plasma-treated hexadecyl monolayers on Si(111) substrates, after different durations of plasma treatment.
Figure 6. Surface reactions during plasma treatment of alkyl monolayers on silicon.
behavior, the Si(111) substrate was not measurably affected by plasma treatment of monolayers. This difference in quality of the monolayers, already observed in their initial water contact angles (109 ( 1° and 111 ( 1° on Si(100) and Si(111) surfaces, respectively) can be explained by the different etching procedures of the two substrates. Indeed, unlike for Si(100), the buffered etching of Si(111) surfaces allows the formation of large atomically flat terraces, ensuring a better saturation of the surface after reaction with 1-hexadecene.52 The resulting monolayer is apparently of such a high quality that surface passivation continues even upon top-layer oxidation by the oxygen plasma, and the Si substrate remains oxygen-free. This unique feature is of specific relevance for electronic applications of the resulting functionalized monolayers, as such applications on semiconductors require minimum surface defects. Plasma-treated samples were also characterized using IRRAS, where the observation of CHn and CdO stretching vibrations corroborates the previous results. As can be seen in Figure 5, the initial intensity of symmetric and antisymmetric CH2 vibrations at 2851 and 2920 cm-1, respectively, gradually decreases upon increasing the duration of the plasma treatment. Meanwhile, the CH3 vibration at 2965 cm-1 even disappears totally after 3 s of treatment, in line with complete loss of methyl-termination and formation of hydrophilic surfaces (water contact angle ≈ 0°). The exact position of 2920 cm-1 for the symmetric CH2 peaks implies a good packing density, since disordered monolayers can display values up to 2926 cm-1.56-58 It is remarkable that the plasma (56) Linford, M. R.; Chidsey, C. E. D. J. Am. Chem. Soc. 1993, 115, 12631. (57) Arafat, A.; Giesbers, M.; Rosso, M.; Sudh€olter, E. J. R.; Schroen, K.; White, R. G.; Yang, L.; Linford, M. R.; Zuilhof, H. Langmuir 2007, 23, 6233–6244. (58) Arafat, A.; Schroen, K.; de Smet, L. C. P. M.; Sudh€olter, E. J. R.; Zuilhof, H. J. Am. Chem. Soc. 2004, 126, 8600–8601.
870 DOI: 10.1021/la9023103
Rosso et al.
treatment does not influence the positions of both CH2 peaks, even as their intensities decrease. This observation shows that the packing of the alkyl monolayers was not significantly affected by the plasma, even upon oxidation of the top surface. While the underlying alkyl chains remain unaffected by the plasma treatment, the growth of shoulders on the high-frequency side of the CH2 peaks clearly reveals the appearance of a less dense layer in the upper part of the monolayer. Oxidation thus starts at the top of the monolayer, during which process no sputtering or desorption of entire alkyl chains is observed, as it would lead to holes in the monolayers and subsequent oxidation of the substrate (see Figure 6 for the case of Si(111) surfaces and Figure S2 of the Supporting Information for the case of Si(100) surfaces).16 Despite a low signal-to-noise ratio, IRRAS also allowed us to monitor the growth of the characteristic peak of CdO stretching vibrations. The low signal intensity achievable with IRRAS and the presence of several carbonyl and carboxyl functionalities on the surface (aldehydes, ketones, and carboxylic acids) resulted in a wide peak positioned at 1715 to 1728 cm-1. The IRRAS data only gives a qualitative confirmation of the presence of CdO functionalities, and more accurate IR data would require the use of other techniques such as ATR.22 Attachment of Au NPs on Plasma-Treated Monolayers and Patterning. For the attachment of gold nanoparticles (Au NPs), we only used Si(111) substrates, as these surfaces could completely withstand the plasma oxidation while allowing further functionalization. After plasma treatment of hexadecyl monolayers for 1 s, the resulting surface aldehydes were reacted with cysteamine (2-aminoethanethiol) to give substrates terminated with thiol functionalities. When samples were then placed in a solution of Au NPs, these would adsorb at the surface and remain attached due to the formation of Au-S bonds. After cleaning and sonication of the substrates in water, the dense coating of nanoparticles could be observed with atomic force microscopy (AFM; see Figure 7, right). The size of ∼15 nm for individual Au NPs can be measured by the height-to-height distance on the AFM section, showing the presence of a single monolayer and the absence of aggregates. For comparison, Figure 7(left) shows the AFM picture of a monolayer-coated Si(111) surface before plasma treatment, which only displays the characteristic terraces of the underlying Si(111) surface. Subsequently, the grafting of Au NPs was used to demonstrate the possibility of surface patterning using plasma. Indeed, the combination of plasma treatments and classical photolithography techniques has already been proposed for the fine patterning of surfaces,57,59,60 but;to the best of our knowledge;the combination of microcontact stamps and plasma treatment has not been applied to monolayer-coated surfaces yet. When a soft-patterned PDMS mask was pressed to the monolayers during the plasma treatment, the oxidation could be fully restricted to the uncovered area of the substrates. The result of such an experiment is presented in the AFM images in Figure 8. The nonexposed areas are comparable to Figure 7(left) and retain the strong hydrophobic character indicative of the persistence of the intact hexadecyl monolayer. The boundary between the different areas displays two specific features: (1) At the edge of the areas that were covered with the PDMS stamp, an elevated line is visible, which did not appear when stamps were applied without plasma treatment. The leakage of low-molecular-weight oligomers of PDMS has already been described in previous surface-patterning (59) Ohl, A.; Schroder, K. Surf. Coat. Technol. 1999, 119, 820–830. (60) Lercel, M. J.; Craighead, H. G.; Parikh, A. N.; Seshadri, K.; Allara, D. L. J. Vac. Sci. Technol., A 1996, 14, 1844–1849.
Langmuir 2010, 26(2), 866–872
Rosso et al.
Article
Figure 7. AFM pictures and sections measured on monolayer-coated silicon(111) surfaces. Left: After exposure to cysteamine and gold nanoparticles without plasma treatment; right: after 1 s of air-based plasma followed by reaction with cysteamine and gold nanoparticles.
Figure 8. Atomic force microscopy (AFM) image of monolayercoated silicon surfaces that are patterned during plasma treatment and subsequently reacted with cysteamine and Au NPs.
experiments involving PDMS stamps,61,62 and we attribute this to stamp edge-induced oxidation. (2) The coverage of Au NPs does not show an abrupt transition but a regular decrease over 500 nm as they approach the boundary. This imperfection is probably due to limitations in the diffusion of the plasma at the edge of the stamp, during the short reaction time. Nevertheless, the observed thin transition area (∼500 nm) allows the formation of complex micrometer-sized patterns, using elastomeric stamps: such a combination of plasma oxidation and PDMS stamps for patterning was already proven successful on polymer63 and glass64 surfaces. For the particular case of organic monolayers, where short reaction times are needed, thin stamps with perforated micrometer-sized features, such as polymeric microsieves,65 would be ideal. Formation of Biospecific Surfaces. To demonstrate the easy biofunctionalization of alkene-based monolayers using plas(61) Jun, Y.; Le, D.; Zhu, X. Y. Langmuir 2002, 18, 3415–3417. (62) Yang, L.; Shirahata, N.; Saini, G.; Zhang, F.; Pei, L.; Asplund, M. C.; Kurth, D. G.; Ariga, K.; Sautter, K.; Nakanishi, T.; Smentkowski, V.; Linford, M. R. Langmuir 2009, 25, 5674–5683. (63) Langowski, B. A.; Uhrich, K. E. Langmuir 2005, 21, 10509–10514. (64) Hwang, H.; Kang, G.; Yeon, J. H.; Nam, Y.; Park, J. K. Lab Chip 2009, 9, 167–170. (65) Girones, M.; Akbarsyah, I. J.; Nijdam, W.; van Rijn, C. J. M.; Jansen, H. V.; Lammertink, R. G. H.; Wessling, M. J. Membr. Sci. 2006, 283, 411–424. (66) Coffinier, Y.; Boukherroub, R.; Wallart, X.; Nys, J. P.; Durand, J. O.; Stievenard, D.; Grandidier, B. Surf. Sci. 2007, 601, 5492–5498.
Langmuir 2010, 26(2), 866–872
ma oxidation, hexadecyl monolayers were prepared on silicon nitride,57,58,66,67 using a hydrosilylation reaction similar to the one used for silicon. In particular, we used a UV-mediated reaction of alkenes with etched silicon nitride surfaces67 and a plasma treatment as described above. The specific surface configuration (monolayer on silicon nitride deposited onto a silicon wafer) allows the use of reflectometry to monitor in situ protein adsorption after attachment of specific biomolecules.53 For this purpose, plasma-treated hexadecyl monolayers were reacted with biotin hydrazide, and the resulting biotin-terminated substrate was incubated with avidin in PBS buffer, and then with BSA. As depicted in Figure 9, avidin still has some of its 4 recognition sites available after a specific adsorption with the immobilized biotin.68 The final saturation of free surface sites with BSA ensures that no protein will adsorb unspecifically during the reflectometry measurements.22 The reflectometry measurements displayed in Figure 10 show the specific interaction of biotin-labeled BSA with avidinfunctionalized silicon nitride surfaces: during the first phase of exposure to unlabeled BSA, no adsorption occurred, because all the areas available for nonspecific adsorption have already been saturated, either by avidin attachment or by the subsequent incubation with BSA. When biotin-labeled BSA is introduced in the cell, an adsorption of 0.5 mg/m2 was measured, due to the specific avidin-biotin interaction. For comparison, reflectometry experiments were also carried out without the final surface saturation with BSA (See Figure S3 of the Supporting Information). This time, nonspecific adsorption of unlabeled BSA (∼1 mg/m2) is observed, prior to the adsorption of biotinylated BSA identical to that observed in the previous experiment. From these adsorbed amounts, we can deduce that the biotin-labeled BSA can still occupy roughly one-third of the available surface: the total surface saturation (BSA and biotin-labeled BSA) of 1.5 mg/m2 was comparable with the normal coverage of BSA on bare oxide surfaces. The surface density of the functionalization obtained with the plasma treatment is an important issue if this method has to compete with classical chemical reactions. It is obvious that the simplicity, speed, and effectiveness make this method highly (67) Rosso, M.; Giesbers, M.; Arafat, A.; Schroen, K.; Zuilhof, H. Langmuir 2009, 25, 2172–2180. (68) Christman, K. L.; Requa, M. V.; Enriquez-Rios, V. D.; Ward, S. C.; Bradley, K. A.; Turner, K. L.; Maynard, H. D. Langmuir 2006, 22, 7444–7450.
DOI: 10.1021/la9023103
871
Article
Rosso et al.
Figure 9. Formation of avidin-terminated surfaces, using plasma treatment, and subsequent reactions with biotin hydrazide and avidin, followed by passivation of the remaining areas with BSA.
nanoparticles and biomolecules can also easily be achieved, while the use of a patterned mask during the plasma treatment results in micrometer-sized features. This is a new technique for the spatially controlled chemical and biochemical functionalization of silicon and silicon nitride surfaces, which can likely also be applied to organic monolayers that are covalently attached onto other surfaces, such as noble metals, oxidic surfaces, or silicon carbide. In particular for glass, it would complement currently existing soft lithographical69 or photochemical70 means to obtain patterned surfaces.
Figure 10. Reflectometry data obtained by subjecting avidincoated silicon nitride surfaces to BSA and biotin-labeled BSA.
competitive, specifically for use of binding of relatively big species, such as the proteins or nanoparticles presented in this work.
Conclusions The reported results show the possibility to use a simple and short (maximally 3 s) plasma treatment to functionalize methylterminated monolayers on silicon and silicon nitride surfaces in a fast and reproducible way. The quality of the monolayers and substrates was not affected by the activation of the top molecular layer of the coatings, as shown with IRRAS and XPS. The plasma treatment allows the ready attachment of functional proteins, as demonstrated by the avidin-biotin interaction. Grafting of gold
872 DOI: 10.1021/la9023103
Acknowledgment. The authors thank Graduate School VLAG and MicroNed (Project no. 6163510395) for financial support and Remko Fokkink (Wageningen University) for technical support. Supporting Information Available: Details of the calculations of adsorbed amounts from reflectometry data, narrowscan XPS C1s, and IRRAS measured on plasma-treated 1-hexadecene monolayers on Si(100) surfaces, and control reflectometry measurement of biotinylated BSA on streptavidin without passivation with unlabeled BSA. This material is available free of charge via the Internet at http://pubs. acs.org. (69) Onclin, S.; Ravoo, B. J.; Reinhoudt, D. Angew. Chem., Int. Ed. 2005, 44, 6282–6304. (70) ter Maat, J.; Regeling, R.; Yang, M.; Mullings, M. N.; Bent, S. F.; Zuilhof, H. Langmuir 2009, accepted; DOI: 10.1021/la901551t.
Langmuir 2010, 26(2), 866–872
